Discovery of 7α-substituted dihydrotestosterones as androgen receptor pure antagonists and their structure-activity relationships

Article abstract of DOI:10.1016/j.bmc.2006.09.072

A series of 7α-substituted dihydrotestosterone derivatives were synthesized and evaluated for androgen receptor (AR) pure antagonistic activity. From reporter gene assay (RGA), the compound with a side chain containing N-n-butyl-N-methyl amide (19a) showed pure antagonistic activity (IC₅₀ = 340 nM, FI₅ > 10,000 nM), whereas known AR antagonists showed partial agonistic activities. The optimization of 19a led to compound 23 (CH4892280), which showed more potent pure antagonistic activity (IC₅₀ = 190 nM, FI₅ > 10,000 nM). The SARs of tested compounds suggested that the length of the side chain and the substituents on the amide nitrogen are important for pure antagonistic activities.

Full text of DOI:10.1016/j.bmc.2006.09.072

Bioorganic & Medicinal Chemistry 15 (2007) 174–185
Discovery of 7a-substituted dihydrotestosterones as androgen
receptor pure antagonists and their structure–activity relationships
Kazutaka Tachibana,* Ikuhiro Imaoka, Hitoshi Yoshino, Takashi Emura,
Hirohumi Kodama, Yoshiyuki Furuta, Nobuaki Kato, Mitsuaki Nakamura,
Masateru Ohta, Kenji Taniguchi, Nobuyuki Ishikura, Masahiro Nagamuta,
Etsuro Onuma and Haruhiko Sato
Fuji Gotemba Research Laboratories, Chugai Pharmaceutical Co., Ltd, 1-135 Komakado, Gotemba, Shizuoka 412-8513, Japan
Received 16 August 2006; revised 28 September 2006; accepted 30 September 2006
Available online 4 October 2006
Abstract—A series of 7a-substituted dihydrotestosterone derivatives were synthesized and evaluated for androgen receptor (AR)
pure antagonistic activity. From reporter gene assay (RGA), the compound with a side chain containing N-n-butyl-N-methyl amide
(19a) showed pure antagonistic activity (IC50 = 340 nM, FI5 > 10,000 nM), whereas known AR antagonists showed partial agonistic
activities. The optimization of 19a led to compound 23 (CH4892280), which showed more potent pure antagonistic activity
(IC50 = 190 nM, FI5 > 10,000 nM). The SARs of tested compounds suggested that the length of the side chain and the substituents
on the amide nitrogen are important for pure antagonistic activities.
ꢀ 2006 Elsevier Ltd. All rights reserved.
1. Introduction
breast cancer11,12 and was launched in 2002. Moreover,
it was found in the X-ray crystallography of the complex
of 5 and ligand-binding domain (LBD) of ERb that the
side chain of 5 inhibits the folding of helix 12 of ERb
LBD,13 presumed to be the important mechanism of
expression of pure antagonistic activity. As AR pure
antagonists, Hashimoto et al. discovered isoxazolone
derivatives designed on the basis of ‘the helix-folding
Prostate cancer is the most common cancer amongst
men in the USA and the second most common malig-
nant cause of male death worldwide after lung cancer.1
Since the growth of prostate cancer is dependent on
androgen, androgen receptor (AR) antagonists such as
flutamide (1) and bicalutamide (3) are currently used
as hormone therapy.2 These antiandrogens exhibit good
efficacy in many cases and comprise an important part
of effective therapeutics.3–6 However, the most consider-
able problem with these antiandrogens is that recurrence
occurs after a short period of response.7 Since they have
partial agonistic activities at high concentration in vitro,8
this may be attributed to recurrence. Therefore, new
antiandrogenic agents that exhibit no agonistic activi-
ties, so-called ‘AR pure antagonists’, are expected.
H
N
H
N
OH
R
O
S
O
O2N
F3C
NC
F3C
F
O
O
flutamide (1)
bicalutamide (3)
(R=H)
hydroxyflutamide (2)
(R=OH)
It has been reported that 7a-substituted estradiols such
as fulvestrant (4)9 and ICI 164,384 (5)10 show pure
antagonistic activities against estrogen receptor (ER).
Fulvestrant showed efficacy in tamoxifen-resistant
OH
HO
R
fulvestrant (4) (R=(CH2)9SO(CH2)3CF2CF3)
ICI164,384 (5) (R=(CH2)10CON(Me)nBu)
Keywords: Androgen receptor; Pure antagonist; Dihydrotestosterone;
Prostate cancer.
*
Chart 1. Structures of AR and ER antagonists.
Corresponding author. E-mail: tachibanakzt@chugai-pharm.co.jp
0968-0896/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2006.09.072

K. Tachibana et al. / Bioorg. Med. Chem. 15 (2007) 174–185
175
inhibition hypothesis’.14,15 Since AR and ER belong to
the nuclear receptor superfamily and the tertiary struc-
ture of AR LBD is similar to that of ER LBD,16,17 we
hypothesized that dihydrotestosterone (DHT) deriva-
tives which have side chain in position 7a similar to
ER pure antagonists could be AR pure antagonists
(see Chart 1).
dures,18–20 was converted to the key intermediate 10 in
four steps. For the next step, we selected olefin cross-me-
tathesis21 as the key reaction to effectively introduce side
chains of various lengths. The reaction of 10 and two
equivalents of bromoalkenes in the presence of a catalyt-
ic amount of Cl2(Cy3P)2Ru@CHPh gave 11a–e in yields
of 69–98%. Further derivatizations to the target com-
pounds were carried out by two routes as described in
Scheme 1. For both routes, bromide 11 was converted
to 4,4,5,5,5-pentafluoropentylsulfides 13 orꢁ14 followed
by oxidation to a sulfoxide using OXONE .22
In this paper, we report the design and synthesis of new
DHT derivatives with the side chain in position 7a and
their AR pure antagonistic activity in vitro. Moreover,
we report the structure–activity relationships deter-
mined from the process of the investigation.
The synthetic route of the amide side chain derivatives
(19a–e, 20–37) is shown in Scheme 2. In this scheme
also, olefin cross-metathesis was adopted as the key
reaction to form alkenyl esters 17a–f. After the reduc-
tion of the olefins by Pd-catalyzed hydrogenation,
removal of the tert-butyldimethylsilyl group and
hydrolysis of the ester were performed simultaneously
by 1 N HCl to give carboxylic acids 18a–f. Finally,
these compounds were converted to target compounds
2. Chemistry
The target compounds with a sulfoxide side chain (16a–
e) were prepared according to the synthetic sequence
outlined in Scheme 1. Starting material 6, synthesized
from testosterone according to the reported proce-
OR
OTBS
OR
e
a
c
O
O
H
O
O
H
6
9 (R=H)
10 (R=MOM)
d
7 (R = TBS)
8 (R = H)
b
OMOM
OMOM
f
O
O
(CH2)m-3Br
(CH2)mBr
O
H
O
H
11a (m=5)
b (m=7)
12a (m=5)
b (m=13)
c (m=9)
d (m=11)
e (m=13)
g
g, h
OH
OMOM
O
(CH2)m-3S(CH2)3CF2CF3
14a (m=7)
O
(CH2)mS(CH2)3CF2CF3
H
i
O
H
13a (m=5)
b (m=13)
b (m=9)
c (m=11)
h, i
OH
OH
f
O
O
(CH2)m-3S(CH2)3CF2CF3
O
O
(CH2)mS(CH2)3CF2CF3
H
H
15a (m=7)
16a~e
b (m=9)
c (m=11)
Scheme 1. Reagents and conditions: (a) Li, liquid NH3, ꢀ78 ꢂC, 20 min; (b) 2 N HCl, acetone, rt, 4 h; (c) ethylene glycol, p-TsOH, benzene, reflux,
2 h; (d) MOMCl, iPr2NEt, CH2Cl2, rt, 14 h; (e) CH2@CH(CH2)mꢀ3Br, Cl2(Cy3P)2Ru@CHPh (cat.), CH2Cl2, reflux; (f) H2, 10% Pd/C (cat.), AcOEt,
rt; (g) AcS(CH2)3CF2CF3, 1 M MeONa in MeOH, THF–MeOH, rt; (h) 2 N HCl, acetone, 60 ꢂC; (i) OXONEꢁ, THF–H2O, 0 ꢂC, 30 min.

176
K. Tachibana et al. / Bioorg. Med. Chem. 15 (2007) 174–185
OTBS
OTBS
b
a
(CH2)m-3COOR
O
O
H
H
7
17a (m=6)
b (m=7)
c (m=8)
d (m=9)
e (m=10)
f (m=11)
OH
OH
c or d or e
O
(CH2)mCONRR'
O
(CH2)mCOOH
H
H
18a (m=6)
b (m=7)
c (m=8)
d (m=9)
e (m=10)
f (m=11)
19a-e, 20~37
Scheme 2. Reagents and conditions: (a) CH2@CH(CH2)mꢀ3CO2R, Cl2(Cy3P)2Ru@CHPh (cat.), CH2Cl2, reflux; (b) H2, 10% Pd/C (cat.), AcOEt, rt,
then 1 N HCl, acetone, reflux; (c) HNRR0, EDC, HOBt, THF, rt; (d) HNRR0, HATU, iPr2NEt, THF, rt; (e) ClCO2Et, NEt3, THF, 0 ꢂC, then NH3
gas.
19a–e and 20–37 by amidation using corresponding
amines.
in Table 1. They have various lengths of side chains
which have the 4,4,5,5,5-pentafluoropentylsulfoxide
group (16a–e) or the N-n-butyl-N-methylamide group
(19a–e). Most of the sulfoxide derivatives exhibited no
antagonistic activities, only 16c showed weak antagonis-
tic activity (IC50 = 2900 nM), and all of them exhibited
agonistic activities. These results were quite different
from those from ER antagonists.
3. Biological assay
Affinities of tested compounds against AR were deter-
mined by competitive receptor-binding assay using
[3H]mibolerone and CHO-K1/hAR cells.8,23 Agonistic
and antagonistic activities of the compounds for AR
were determined by reporter gene assay (RGA) using
hAR-transfected HeLa cells.24,25 Antagonistic activity
was described as IC50 value, which is the concentration
of a compound to inhibit the transcriptional activity of
0.1 nM of DHT by 50%. To describe agonistic activity,
we calculated the value of ‘FI5’, the concentration of a
compound-treated group in which the transcriptional
activity is five times the transcriptional activity of the
case without the addition of a compound. A ‘pure
antagonist’ was defined to have a FI5 value greater than
10,000 nM.
In the case of N-n-butyl-N-methylamide derivatives,
their profiles were dependent on the length of the side
chain. Specifically, 19b (m = 8), 19d (m = 10), and 19e
(m = 11) have agonistic activities, but 19a (m = 7) and
19c (m = 9) showed no agonistic activities even at
10,000 nM. Therefore, by our definition, 19a and c are
considered to be the pure antagonists.
In the next step, we attempted the optimization of the
N-substituents of 19a (Table 2).
The same activities as 19a were retained by the substitu-
tion of an n-butyl group with smaller groups such as
with isopropyl (21), ethyl (22), or methyl (23). The
antagonistic activity was decreased by the substitution
of an n-butyl group to a larger group such as a benzyl
group (25). Moreover, the RGA activity and binding
affinity disappeared when both groups on the nitrogen
atom were substituted for larger groups (26). On the
other hand, although agonistic activity resulted from
removal of substituents from the nitrogen atom (30),
the introduction of a methyl group also led to the
disappearance of agonistic activity (31). The larger the
methyl group of 31 was, the lower the antagonistic
activities became (32–35). The compounds with cyclic
4. Results
The characterization of hydroxyflutamide (2), an active
metabolite of flutamide, and bicalutamide (3) in RGA
(Table 1) showed antagonistic activities of IC50s of 31
and 140 nM, and FI5 values of 1000 and 770 nM,
respectively, and thus clarified that these known AR
antagonists have partial agonistic activities.
As the first step in our investigation, we designed and
synthesized the compounds (16a–e and 19a–e) shown

K. Tachibana et al. / Bioorg. Med. Chem. 15 (2007) 174–185
177
Table 1. Binding and agonistic/antagonistic activities of 7a-substituted DHT derivativesa
OH
O
R
Compound
R
m
Binding affinity IC50 (nM)
RGA
FI5 (nM)
IC50 (nM)
16a
16b
(CH2)m–SO–(CH2)3C2F5
(CH2)m–SO–(CH2)3C2F5
(CH2)m–SO–(CH2)3C2F5
(CH2)m–SO–(CH2)3C2F5
(CH2)m–SO–(CH2)3C2F5
(CH2)m–CON(Me)nBu
(CH2)m–CON(Me)nBu
(CH2)m–CON(Me)nBu
(CH2)m–CON(Me)nBu
(CH2)m–CON(Me)nBu
—
5
6,800
960
110
16
200
>10,000
>10,000
2,900
>10,000
>10,000
340
7
9
16c
16d
490
120
11
13
7
>10,000
>10,000
280
16e
19a
310
>10,000
41
19b
19c
8
550
570
3,400
1,800
1,300
3,500
31
9
>10,000
31
410
19d
19e
10
11
—
—
1,900
9,900
200
2 (hydroxyflutamide)
3 (bicalutamide)
1,000
770
—
200
140
a All data are mean values of duplicate experiments.
Table 2. Binding and agonistic/antagonistic activities of 19a derivativesa
OH
O
(CH2)m
O
H
NRR'
Compound
m
NRR0
Binding affinity IC50 (nM)
RGA
FI5 (nM)
IC50 (nM)
19a
20
21
22
23
24
25
26
7
7
7
7
7
7
7
7
N(Me)nBu
N(Me)Pr
N(Me)iPr
N(Me)Et
NMe2
280
300
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
340
1,100
420
430
660
420
190
260
120
NEt2
N(Me)Bn
N(n-hexyl)2
300
1,600
>10,000
820
>10,000
27
28
29
7
7
7
1,300
350
>10,000
>10,000
>10,000
730
300
490
N
N
120
30
31
32
33
34
35
36
37
7
7
7
7
7
7
6
6
NH2
NHMe
NHEt
NHnPr
NHnBu
NHn-hexyl
NMe2
260
710
2,400
>10,000
>10,000
>10,000
1,600
450
480
1,500
1,600
NTb
2,100
110
940
1,800
6,300
6,600
200
>10,000
490
1,700
NEt2
330
530
a All data are mean values of duplicate experiments unless otherwise noted.
b Not tested.

178
K. Tachibana et al. / Bioorg. Med. Chem. 15 (2007) 174–185
A
B
100
75
50
25
0
20
15
10
5
23
2
23
2
3
3
0
1
10
100
1000 10000
1
10
100
1000 10000
concentration (nM)
concentration (nM)
Figure 1. Agonistic and antagonistic activities of CH4892280 (23), hydroxyflutamide (2), and bicalutamide (3) in reporter gene assay with hAR-
transfected HeLa cells. (A) Dose-dependent agonistic activities of compounds without DHT. (B) Dose-dependent antagonistic activities of
compounds in the presence of 0.1 nM of DHT. All data are mean values of duplicate experiments.
groups (27–29) as well as 19a exhibited pure antagonistic
activities.
over the wide range of 11–19 atoms in length. However,
contrary to our expectations, these compounds exhibit-
ed full agonistic activities with FI5 values of 16–
486 nM, with the exception of 16c, which showed weak
antagonistic activity. That the side chains of these com-
pounds may not prevent the folding of helix 12 to the
agonistic form can be explained by the following two
reasons: (1) the side chain might turn in a direction in
which it cannot inhibit the folding of helix 12 to an ago-
nistic form and (2) compounds such as 16d and 16e
might have exhibited agonistic activities by the allosteric
effect because they showed no binding affinities to AR,
even though it is extraordinary that there are such differ-
ences in the binding affinities of the compounds in which
there are only slight differences in the length of side
chain (e.g., 16c vs 16d).
With the aim of optimization of the length of side chain,
two compounds with side chains shorter by one carbon
than the others were synthesized (36 and 37). However,
both of them showed only partial agonistic activities.
Compound 23 (CH4892280) exhibited the highest antag-
onistic activity of the pure antagonists.
The dose-dependent agonistic and antagonistic activities
of CH4892280 (23), hydroxyflutamide (2), and bicaluta-
mide (3) in RGA are shown in Figure 1. CH4892280
(23) showed no agonistic activity even at 10,000 nM,
but hydroxyflutamide (2) and bicalutamide (3) exhibited
obvious agonistic activities at 100 nM which increased
dose dependently (Fig. 1A). On the other hand,
CH4892280, but not 2 or 3, inhibited transcriptional
activity of 0.1 nM DHT completely at 10,000 nM
(Fig. 1B).
For N-methyl-N-n-butyl amide derivatives, contrary to
the sulfoxide derivatives, the agonistic activities disap-
peared alternately in the range of seven to 10 methylene
groups (19a and 19c). Therefore, compounds with a sulf-
oxide side chain which have an even number of methyl-
ene groups might exhibit pure antagonistic activity.
5. Discussion
In the optimization of 19a, it was found that the antag-
onistic activities are influenced by substituents on the
nitrogen atom of the amide moiety. In the case of
N,N-disubstituted derivatives, the same level of antago-
nistic activities was expressed in the N,N-dimethyl deriv-
ative (23) through the N-n-butyl-N-methyl derivative
(19a). On the contrary, with N-monosubstituted deriva-
tives, antagonistic activities decreased as the substituent
became larger. These distinct results were probably the
result of the difference in the conformation of the sub-
stituents; that is, in N,N-disubstituted amides, the sub-
stituents may have turned to a direction different than
the substituent in N-monosubstituted amides because
of the steric hindrance of the geminal substituent.
The value ‘FI5’ was established as the criterion for ago-
nistic activity of a compound. Hydroxyflutamide (2)
exhibited an FI5 of 1000 nM in this study and the ago-
nistic activity was comparable to the concentration as
reported in the literature.23 Plasma levels of 2 in prostate
cancer patients have been reported to reach 78 ng/ml
(8 lM).23 Therefore, we considered our criterion to be
valid.
Our strategy for the discovery of AR pure antagonists
was based on the fact that estradiol derivatives having
a side chain in position 7a exhibit pure antagonistic
activities against ER. This suggests that the folding of
helix 12 of ER LBD is inhibited by the side chain of
the antagonist sticking out of the pocket.13 It has also
been reported that the optimal length of the side chain
is 16–18 atoms.10 Since the optimal length of the side
chain has not been clarified for AR pure antagonists,
we synthesized compounds with sulfoxide side chains
We also studied the effect of CH4892280 (23) on seminal
vesicle wet weight in mice to investigate the in vivo anti-
androgenic activity of the compound. However,
CH4892280 showed no activities from either by oral or
subcutaneous administration (data not shown). These

K. Tachibana et al. / Bioorg. Med. Chem. 15 (2007) 174–185
179
results might be attributed to the metabolic instability of
the compound because it was significantly instable in rat
liver microsome (data not shown). Improvement of met-
abolic stability would be important for the clarification
of the in vivo activities of AR pure antagonists.
white solid. Mp 197–198 ꢂC; 1H NMR (270 MHz,
CDCl3) d 0.77 (3H, s), 1.04 (3H, s), 0.96–2.44 (23H,
m), 3.60–3.70 (1H, m), 4.94 (1H, d, J = 3.5 Hz), 5.00
(1H, s), 5.58–5.72 (1H, m); Rf 0.54 (AcOEt/hexane
1:1). Anal. Calcd for C22H34O2: C, 79.95; H, 10.37.
Found: C, 79.73; H, 10.39.
6. Conclusion
7.1.3. 7a-Allyl-3,3-ethylenedioxy-17b-hydroxy-5a-andro-
stane (9). To a solution of 8 (126 mg, 0.381 mmol) in
benzene (5 ml) were added ethylene glycol (2 ml) and
p-TsOH (13.2 mg, 0.0767 mmol), and the mixture was
refluxed for 2 h using Dean-Stark trap. After cooling
to room temperature, satd NaHCO3 aq was added
under cooling with ice and extracted with AcOEt. The
organic layer was washed with brine and dried over
MgSO4, filtered, and concentrated in vacuo to give 9
(141 mg, 99%). 1H NMR (270 MHz, CDCl3) d 0.74
(3H, s), 0.85 (3H, s), 0.92–2.20 (23H, m), 3.56–3.70
(1H, m), 3.93 (4H, s), 4.92–5.04 (2H, m), 5.62–5.80
(1H, m); Rf 0.60 (AcOEt/hexane 1:1). This material
was used in the next step without purification.
Some of the compounds synthesized in this study includ-
ing CH4892280 (23) exhibited AR antagonistic activities
of IC50s of 100 nM order and no agonistic activities even
at 10,000 nM and thus meet the criteria for AR pure
antagonists in vitro. Although the structure–activity
relationships (e.g., optimum length or optimum struc-
ture of side chain) of the compounds differ, our hypoth-
esis that AR pure antagonists could be obtained by the
introduction of a side chain to the natural ligand, similar
to ER pure antagonists, have been established.
7. Experimental
7.1.4. 7a-Allyl-3,3-ethylenedioxy-17b-methoxymethoxy-
5a-androstane (10). To a solution of 9 (141 mg,
0.376 mmol) in CH2Cl2 (4 ml) were added iPr2NEt
(0.227 ml, 1.33 mmol) and MOMCl (0.087 ml,
1.16 mmol) dropwise under cooling with ice. After stir-
ring for 14 h at 25 ꢂC, satd NaHCO3 aq was added to
the reaction mixture and extracted with CH2Cl2. The
organic layer was washed with brine and dried over
MgSO4, filtered, and concentrated in vacuo. The residue
was purified by column chromatography (AcOEt/hex-
ane 1:4) to give 10 (131 mg, 83%) as white solid. Mp
7.1. Chemistry: instruments
Column chromatography was carried out on Merck Sil-
icagel 60 (230–400 mesh) if not otherwise specified. Rf
was determined with Merck Silicagel 60 F254 plates.
1H NMR spectra were recorded on JEOL EX-270. Mass
spectra (MS) were measured with VG Analytical VG70-
250SEQ (FAB) or Thermo Electron LCQ Classic (ESI).
High resonance mass spectra (HRMS) were recorded by
a Micromass Q-Tof Ultima API mass spectrometer.
1
112–114 ꢂC; H NMR (270 MHz, CDCl3) d 0.77 (3H,
s), 0.84 (3H, s), 0.92–2.20 (23H, m), 3.35 (3H, s), 3.53
(1H, t, J = 8.3 Hz), 3.92 (4H, s), 4.62 (2H, d,
J = 1.8 Hz), 4.92–5.04 (2H, m), 5.62–5.80 (1H, m); Rf
0.50 (AcOEt/hexane 1:4). Anal. Calcd for C26H42O4:
C, 74.60; H, 10.11. Found: C, 74.40; H, 10.13.
7.1.1. 7a-Allyl-17b-(tert-butyldimethylsilyl)oxy-5a-andro-
stan-3-one (7). Li (220 mg, 31.7 mmol) was added to
liquid ammonia (150 ml) at ꢀ78 ꢂC. After stirring for
5 min,
7a-allyl-17b-[(tert-butyldimethylsilyl)oxy]and-
rost-4-en-3-one (6) (1.261 g, 2.85 mmol) and tert-buta-
nol (0.41 ml) in THF (20 ml) were added and the
mixture was stirred for 20 min. 1,2-Dibromoethane
(3 ml) and NH4Cl (30 g) were added, and the mixture
was stirred at 25 ꢂC for 30 min. Water was added to
the mixture and extracted with AcOEt. The organic
layer was dried over MgSO4, filtered, and concentrated
in vacuo. The residue was purified by column chroma-
tography (AcOEt/hexane 1:10) to give 7 (810.7 mg,
64%) as white solid. Mp 130–132 ꢂC; 1H NMR
(270 MHz, CDCl3) d 0.01 (6H, s), 0.73 (3H, s), 0.88
(9H, s), 1.04 (3H, s), 0.92–2.45 (23H, m), 3.55 (1H, t,
J = 8.3 Hz), 4.93 (1H, d, J = 3.8 Hz), 4.99 (1H, s),
5.58–5.72 (1H, m); Rf 0.54 (AcOEt/hexane 1:10).
7.1.5. 7a-(13-Bromo-2-tridecen-1-yl)-3,3-ethylenedioxy-
17b-methoxymethoxy-5a-androstane (11e). The mixture
of 10 (42.6 mg, 0.102 mmol), 12-bromododecene
(50.4 mg, 0.204 mmol), and Cl2(Cy3P)2Ru@CHPh
(8.4 mg, 0.0102 mmol) in CH2Cl2 (1.5 ml) was heated
under reflux for 5 h under Ar atmosphere. After cooling,
purification by column chromatography (AcOEt/hexane
1
1:10) gave 11e (56.0 mg, 86%) as brown oil. H NMR
(270 MHz, CDCl3) d 0.76 (3H, s), 0.83 (3H, s), 0.94–
2.14 (41H, m), 3.34 (3H, s), 3.41 (2H, t, J = 6.9 Hz),
3.52 (1H, t, J = 8.3 Hz), 3.92 (4H, s), 4.62 (2H, d,
J = 1.8 Hz), 5.22–5.46 (2H, m); Rf 0.56 (AcOEt/hexane
1:4).
7.1.2. 7a-Allyl-17b-hydroxy-5a-androstan-3-one (8).
Compound 7 (1.00 g, 2.26 mmol) was dissolved in ace-
tone (20 ml) and then 2 N HCl (2 ml) was added drop-
wise. After stirring for 17 h at room temperature,
water was added to the reaction mixture and extracted
with CH2Cl2. The organic layer was washed with brine,
dried over MgSO4, filtered, and concentrated in vacuo.
The residue was purified by column chromatography
(0–33% AcOEt/hexane) to give 8 (655.7 mg, 88%) as
7.1.6. 7a-(5-Bromo-2-penten-1-yl)-3,3-ethylenedioxy-17b-
methoxymethoxy-5a-androstane (11a). This compound
was prepared using a procedure similar to that described
for 11e. Yield: 79%; 1H NMR (270 MHz, CDCl3) d 0.76
(3H, s), 0.83 (3H, s), 0.90–2.38 (23H, m), 2.50–2.68 (2H,
m), 3.34 (3H, s), 3.37 (2H, t, J = 7.3 Hz), 3.52 (1H, t,
J = 8.3 Hz), 3.92 (4H, s), 4.62 (2H, d, J = 1.8 Hz),
5.30–5.46 (2H, m); Rf 0.50 (AcOEt/hexane 1:4).

180
K. Tachibana et al. / Bioorg. Med. Chem. 15 (2007) 174–185
7.1.7. 7a-(7-Bromo-2-hepten-1-yl)-3,3-ethylenedioxy-17b-
methoxymethoxy-5a-androstane (11b). This compound
was prepared using a procedure similar to that described
for 11e. Yield: 74%; 1H NMR (270 MHz, CDCl3) d 0.77
(3H, s), 0.84 (3H, s), 0.93–2.32 (29H, m), 3.35 (3H, s),
3.40 (2H, t, J = 6.8 Hz), 3.52 (1H, t, J = 8.3 Hz), 3.93
(4H, s), 4.62 (2H, d, J = 1.5 Hz), 5.25–5.46 (2H, m); Rf
0.53 (AcOEt/hexane 1:4); HRMS calcd for C30H50O4Br
553.2892. Found 553.2868.
(2H, t, J = 7.3 Hz), 2.59 (2H, t, J = 7.9 Hz), 3.58–3.70
(1H, m); Rf 0.32 (AcOEt/hexane 1:4).
7.1.12. 17b-Hydroxy-7a-{13-(4,4,5,5,5-pentafluoropentyl-
sulfinyl)tridecyl}-5a-androstan-3-one (16e). To a solution
of 13b (26.0 mg, 0.0391 mmol) in THF (1 ml) were add-
ed OXONEꢁ (14.4 mg, 0.0234 mmol) and water (0.2 ml)
at 0 ꢂC. After stirring for 30 min, satd NaHCO3 aq was
added to the reaction mixture and extracted with
AcOEt. The organic layer was washed with brine and
dried with MgSO4. After filtering, the solvent was dis-
tilled off at reduced pressure. Purification by silica gel
column chromatography (AcOEt/hexane 2:1) gave 16e
(19.5 mg, 73%) as colorless oil. 1H NMR (270 MHz,
CDCl3) d 0.76 (3H, s), 1.04 (3H, s), 0.98–2.40 (50H,
m), 2.58–2.82 (4H, m), 3.60–3.70 (1H, m); MS (FAB)
m/z 681 [(M+H)+]; HRMS calcd for C37H62F5O3S
681.4340. Found 681.4355; Rf 0.10 (AcOEt/hexane 1:4).
7.1.8. 7a-(9-Bromo-2-nonen-1-yl)-3,3-ethylenedioxy-17b-
methoxymethoxy-5a-androstane (11c). This compound
was prepared using a procedure similar to that described
for 11e. Yield: 69%; 1H NMR (270 MHz, CDCl3) d 0.77
(3H, s), 0.83 (3H, s), 0.92–2.30 (33H, m), 3.34 (3H, s),
3.40 (2H, t, J = 6.8 Hz), 3.52 (1H, t, J = 8.2 Hz), 3.92
(4H, s), 4.62 (2H, d, J = 1.6 Hz), 5.18–5.43 (2H, m);
HRMS calcd for C32H54O4Br 581.3205. Found
581.3227.
7.1.13. 17b-Hydroxy-7a-{5-(4,4,5,5,5-pentafluoropentyl-
sulfinyl)pentyl}-5a-androstan-3-one (16a). This com-
pound was prepared from 11a using a procedure similar
to that described for 16e. Colorless oil. 1H NMR
(270 MHz, CDCl3) d 0.76 (3H, s), 1.04 (3H, s), 0.98–
2.40 (34H, m), 2.58–2.82 (4H, m), 3.60–3.70 (1H, m);
MS (FAB) m/z 569 [(M+H)+]; HRMS calcd for
C29H46O3F5S 569.3088. Found 569.3116.
7.1.9. 7a-(11-Bromo-2-undecen-1-yl)-3,3-ethylenedioxy-
17b-methoxymethoxy-5a-androstane (11d). This com-
pound was prepared using a procedure similar to that
described for 11e. Yield: 98%; 1H NMR (270 MHz,
CDCl3) d 0.77 (3H, s), 0.84 (3H, s), 0.95–2.32 (37H,
m), 3.35 (3H, s), 3.41 (2H, t, J = 6.9 Hz), 3.52 (1H, t,
J = 8.2 Hz), 3.92 (4H, s), 4.62 (2H, d, J = 1.7 Hz),
5.20–5.45 (2H, m).
7.1.14. 7a-[7-(4,4,5,5,5-Pentafluoropentylsulfanyl)hept-2-
enyl]-3,3-ethylenedioxy-17b-methoxymethoxy-5a-andro-
stane (14a). 4,4,5,5,5-Pentafluoropentanethioacetate
(29.0 mg, 0.123 mmol) was dissolved in MeOH (1 ml)
and then 1 M MeONa in MeOH (0.115 ml) was added
dropwise. After stirring for 15 min, a solution of 11b
(45.5 mg, 0.0822 mmol) in THF (0.5 ml) was added to
the reaction mixture. After stirring for 13 h, water was
added to the reaction mixture and extracted with
AcOEt. The organic layer was washed with brine and
dried over MgSO4. After filtering, the solvent was dis-
tilled off at reduced pressure. Purification by silica gel
column chromatography (AcOEt/hexane 1:10) gave
14a (49.8 mg, 90%) as colorless oil. 1H NMR
(270 MHz, CDCl3) d 0.77 (3H, s), 0.83 (3H, s), 0.93–
2.27 (33H, m), 2.48–2.61 (4H, m), 3.35 (3H, s), 3.52
(1H, t, J = 8.1 Hz), 3.93 (4 H, s), 4.62 (2H, d,
J = 1.7 Hz), 5.23–5.44 (2H, m); HRMS calcd for
C35H59NO4F5S 684.4085. Found 684.4108 [(M+NH4)+].
7.1.10. 7a-(13-Bromotridecyl)-3,3-ethylenedioxy-17b-meth-
oxymethoxy-5a-androstane (12b). To a solution of 11e
(55.3 mg, 0.0867 mmol) in AcOEt (2 ml) was added
10% Pd/C (10 mg), and the mixture was stirred for
13 h at room temperature in a hydrogen atmosphere.
After filtering the reaction mixture, the solvent was dis-
tilled off at reduced pressure to give 12b (47.4 mg, 86%)
as colorless oil. 1H NMR (270 MHz, CDCl3) d 0.76 (3H,
s), 0.84 (3H, s), 0.94–2.10 (45H, m), 3.34 (3H, s), 3.41
(2H, t, J = 6.9 Hz), 3.53 (1H, t, J = 8.3 Hz), 3.93 (4H,
s), 4.62 (2H, d, J = 1.8 Hz); Rf 0.56 (AcOEt/hexane 1:4).
7.1.11. 17b-Hydroxy-7a-{13-(4,4,5,5,5-pentafluoropentyl-
sulfanyl)tridecyl}-5a-androstan-3-one (13b). 4,4,5,5,5-
Pentafluoropentanethioacetate (35.0 mg, 0.148 mmol)
was dissolved in MeOH (1 ml) and then 1 M MeONa
in MeOH (0.12 ml) was added dropwise. After stirring
for 30 min, a solution of 12b (47.4 mg, 0.0742 mmol)
in THF (1 ml) was added to the reaction mixture. After
stirring for 18 h, water was added to the reaction mix-
ture and extracted with AcOEt. The organic layer was
washed with brine and dried over MgSO4. After filter-
ing, the solvent was distilled off at reduced pressure.
The obtained residue was dissolved in acetone (2 ml),
and then 2 N HCl (0.5 ml) was added. The mixture
was heated under reflux for 3 h at 60 ꢂC. After cooling
to 0 ꢂC, water was added and extracted with CHCl3.
The organic layer was washed with brine and dried with
MgSO4. After filtering, the solvent was distilled off at re-
duced pressure. Purification by silica gel column chro-
matography (AcOEt/hexane 1:4) gave 13b (40.2 mg,
7.1.15. 17b-Hydroxy-7a-{7-(4,4,5,5,5-pentafluoropentyl-
sulfinyl)hept-2-enyl}-5a-androstan-3-one (15a). To
a
solution of 14a (24.5 mg, 0.0367 mmol) in acetone
(2.0 ml) was added 2 N HCl (0.5 ml) and the mixture
was stirred at 60 ꢂC for 2 h. After cooling to room tem-
perature, water was added to the reaction mixture and
extracted with CH2Cl2. The organic layer was washed
with brine and dried over MgSO4. After filtering, the
solvent was distilled off at reduced pressure. The result-
ing residue was dissolved to THF (1 ml) and then
OXONEꢁ (16.6 mg, 0.0271 mmol) and water (0.2 ml)
were added at 0 ꢂC. After stirring for 40 min, satd NaH-
CO3 aq was added to the reaction mixture and extracted
with AcOEt. The organic layer was washed with brine
and dried with MgSO4. After filtering, the solvent was
1
82%) as colorless oil. H NMR (270 MHz, CDCl3) d
0.76 (3H, s), 1.04 (3H, s), 0.88–2.40 (49H, m), 2.50

K. Tachibana et al. / Bioorg. Med. Chem. 15 (2007) 174–185
181
distilled off at reduced pressure. Purification by silica gel
column chromatography (AcOEt) gave 15a (17.6 mg,
66%) as colorless oil. H NMR (270 MHz, CDCl3) d
(1H, t, J = 8.4 Hz), 3.66 (3H, s), 5.20–5.45 (2H, m); Rf
0.42 (AcOEt/hexane 1:15).
1
0.76 (3H, s), 0.95–2.45 (34H, m), 1.04 (3H, s), 2.60–
2.82 (4H, m), 3.63 (1H, brs), 5.20–5.43 (2H, m); Rf
0.31 (AcOEt); HRMS calcd for C31H48O3F5S
595.3244. Found 595.3244.
7.1.21. 9-(17b-tert-Butyldimethylsilyloxy-5a-androstan-
3-one-7a-yl)non-7-enoic acid tert-butyl ester (17c). This
compound was prepared from 7 and tert-butyl 7-octeno-
ate using a procedure similar to that described for 17b.
1
Yield: 74%. Brown oil. H NMR (270 MHz, CDCl3) d
0.01 (6H, s), 0.71 (3H, s), 0.88 (9H, s), 1.03 (3H, s),
0.90–2.08 (37H, m), 2.15–2.42 (5H, m), 3.52 (1H, t,
J = 8.4 Hz), 5.15–5.40 (2H, m); Rf 0.56 (AcOEt/hexane
1:4).
7.1.16. 17b-Hydroxy-7a-{7-(4,4,5,5,5-pentafluoropentyl-
sulfinyl)heptyl}-5a-androstan-3-one (16b). To a solution
of 15a (15.4 mg, 0.0259 mmol) in AcOEt (2 ml) was add-
ed 10% Pd/C (5.0 mg), and the mixture was stirred for
17 h at room temperature in a hydrogen atmosphere.
After filtering the reaction mixture, the solvent was
distilled off at reduced pressure. Purification by silica
gel column chromatography (AcOEt) gave 16b
(13.7 mg, 89%) as colorless oil. 1H NMR (270 MHz,
CDCl3) d 0.76 (3H, s), 1.04 (3H, s), 0.98–2.40 (38H,
m), 2.60–2.82 (4H, m), 3.60–3.70 (1H, m); MS (FAB)
m/z 597 [(M+H)+]; HRMS calcd for C31H50O3F5S
597.3401. Found 597.3402.
7.1.22. 10-(17b-tert-Butyldimethylsilyloxy-5a-androstan-
3-one-7a-yl)dec-8-enoic acid methyl ester (17d). This
compound was prepared from 7 and methyl 8-noneno-
ate using a procedure similar to that described for 17b.
1
Yield: 61%. Brown oil. H NMR (270 MHz, CDCl3) d
0.01 (6H, s), 0.71 (3H, s), 0.87 (9H, s), 1.02 (3H, s),
0.90–2.08 (30H, m), 2.18–2.43 (5H, m), 3.54 (1H, t,
J = 8.3 Hz), 3.66 (3H, s), 5.17–5.41 (2H, m); Rf 0.33
(ether/hexane 1:2).
7.1.17. 17b-Hydroxy-7a-{9-(4,4,5,5,5-pentafluoropentyl-
sulfinyl)nonyl}-5a-androstan-3-one (16c). This compound
was prepared from 11c using a procedure similar to that
7.1.23. 11-(17b-tert-Butyldimethylsilyloxy-5a-androstan-
3-one-7a-yl)undec-9-enoic acid methyl ester (17e). This
compound was prepared from 7 and methyl 9-decenoate
using a procedure similar to that described for 17b.
1
described for 16b. Colorless oil. H NMR (270 MHz,
CDCl3) d 0.76 (3H, s), 1.04 (3H, s), 0.98–2.40 (42H,
m), 2.60–2.82 (4H, m), 3.60–3.70 (1H, m); MS (FAB)
m/z 625 [(M+H)+]; HRMS calcd for C33H54O3F5S
625.3714. Found 625.3708.
1
Yield: 51%. Brown oil. H NMR (270 MHz, CDCl3) d
0.01 (6H, s), 0.71 (3H, s), 0.87 (9H, s), 1.02 (3H, s),
0.90–2.08 (32H, m), 2.17–2.43 (5H, m), 3.53 (1H, t,
J = 8.2 Hz), 3.65 (3H, s), 5.17–5.38 (2H, m); Rf 0.21
(AcOEt/hexane 1:15).
7.1.18. 17b-Hydroxy-7a-{11-(4,4,5,5,5-pentafluoropentyl-
sulfinyl)undecyl}-5a-androstan-3-one (16d). This com-
pound was prepared from 11d using a procedure similar
to that described for 16b. Colorless oil. 1H NMR
(270 MHz, CDCl3) d 0.76 (3H, s), 1.04 (3H, s), 0.98–
2.40 (42H, m), 2.60–2.82 (4H, m), 3.60–3.70 (1H, m);
MS (FAB) m/z 653 [(M+H)+]; HRMS calcd for
C35H58O3F5S 653.4027. Found 653.4036.
7.1.24. 12-(17b-tert-Butyldimethylsilyloxy-5a-androstan-
3-one-7a-yl)dodec-10-enoic acid methyl ester (17f). This
compound was prepared from 7 and methyl 10-undece-
noate using a procedure similar to that described for
17b. Yield: 67%. Brown oil. 1H NMR (270 MHz,
CDCl3) d 0.01 (6H, s), 0.72 (3H, s), 0.88 (9H, s), 1.03
(3H, s), 0.90–2.08 (34H, m), 2.18–2.43 (5H, m), 3.54
(1H, t, J = 8.1 Hz), 3.66 (3H, s), 5.17–5.40 (2H, m); Rf
0.42 (AcOEt/hexane 1:4).
7.1.19. 8-(17b-tert-Butyldimethylsilyloxy-5a-androstan-
3-one-7a-yl)oct-6-enoic acid methyl ester (17b). To a
solution of 7 (596.1 mg, 1.34 mmol) and methyl 6-hept-
enoate (384.4 mg, 2.70 mmol) in CH2Cl2 (5 ml) was
7.1.25. 8-(17b-tert-Butyldimethylsilyloxy-5a-androstan-
3-one-7a-yl)octanoic acid (18b). To a solution of 17b
(505.5 mg, 0.904 mmol) in AcOEt (30 ml) was added
10% Pd/C (148 mg), and the mixture was stirred at room
temperature for 4 h in a hydrogen atmosphere. The reac-
tion mixture was filtered and the solvent was distilled off
at reduced pressure. The resulting residue was dissolved
in acetone (10 ml) and, after adding 1 N HCl (1 ml), the
mixture was heated under reflux for 26 h. After cooling
to room temperature, water was added and extracted
with AcOEt. The organic layer was washed with brine,
dried over MgSO4, filtered, and concentrated at reduced
pressure. Purification by silica gel column chromatogra-
phy (AcOEt/hexane 1:2–1:1) gave 18b (362.6 mg, 93%)
as yellow oil. 1H NMR (270 MHz, CDCl3) d 0.76
(3H, s), 1.04 (3H, s), 1.00–1.82 (27H, m), 1.98–2.15
(3H, m), 2.23–2.48 (5H, m), 3.65 (1H, t, J = 8.7 Hz);
MS (FAB) m/z 433 [(M+H)+]; Rf 0.27 (AcOEt/hexane
1:1).
added
benzylidenebis(tricyclohexylphosphine)dichlo-
roruthenium (57.0 mg, 0.0693 mmol), and the mixture
was heated under reflux for 5 h under Ar. After cooling
to room temperature, the reaction mixture was purified
directly by silica gel column chromatography (AcOEt/
hexane 1:10) to give 17b (527.6 mg, 70%) as brown oil;
1H NMR (270 MHz, CDCl3) d 0.01 (6H, s), 0.71 (3H,
s), 0.88 (9H, s), 1.03 (3H, s), 0.90–2.10 (26H, m), 2.18–
2.43 (5H, m), 3.51 (1H, t, J = 8.4 Hz), 3.67 (3H, s),
5.18–5.40 (2H, m); Rf 0.43 (AcOEt/hexane 1:4).
7.1.20. 7-(17b-tert-Butyldimethylsilyloxy-5a-androstan-
3-one-7a-yl)hept-5-enoic acid tert-butyl ester (17a). This
compound was prepared from 7 and tert-butyl 5-hexe-
noate using a procedure similar to that described for
17b. Yield: 85%. Brown oil. 1H NMR (270 MHz,
CDCl3) d 0.01 (6H, s), 0.71 (3H, s), 0.87 (9H, s), 1.02
(3H, s), 0.90–2.12 (24H, m), 2.18–2.44 (5H, m), 3.54

182
K. Tachibana et al. / Bioorg. Med. Chem. 15 (2007) 174–185
7.1.26. N,N-Dimethyl-8-(17b-hydroxy-5a-androstan-3-
one-7a-yl)octanamide (23). To solution of 18b
7.1.31. N-n-Butyl-N-methyl-12-(17b-hydroxy-5a-andro-
stan-3-one-7a-yl)dodecanamide (19e). This compound
was prepared from 17f using a procedure similar to that
described for 23. Colorless oil. 1H NMR (270 MHz,
CDCl3) d 0.76 (3H, s), 0.92 (3/2H, t, J = 7.4 Hz), 0.95
(3/2H, t, J = 7.4 Hz), 1.04 (3H, s), 1.12–2.45 (46H, m),
2.90 (3/2H, s), 2.96 (3/2H, s), 3.25 (2/2H, t,
J = 7.4 Hz), 3.35 (2/2H, t, J = 7.4 Hz), 3.62 (1H, br);
MS (ESI) m/z 558 [(M+H)+]; HRMS calcd for
C36H64NO3 558.4886. Found 558.4893.
a
(9.9 mg, 0.0229 mmol) in THF (0.5 ml) were added
EDC (13.0 mg, 0.0678 mmol), HOBt (10.5 mg,
0.0686 mmol), and 2.0 M dimethylamine in THF
(68.7 ll, 0.137 mmol). The mixture was stirred at room
temperature for 15 h. After adding AcOEt (2.0 ml), the
mixture was washed with 1 N HCl, satd NaHCO3 aq,
and brine. After being dried over MgSO4, the mixture
was filtered through NH silica gel (DM1020, Fuji Silicia
Chemical Co., Ltd) and the solvent was distilled off at
reduced pressure to give 23 (10.5 mg, 99.7%). 1H
NMR (270 MHz, CDCl3) d 0.76 (3H, s), 1.04 (3H, s),
1.00–1.83 (28H, m), 1.95–2.16 (3H, m), 2.23–2.47 (5H,
m), 2.94 (3H, s), 3.01 (3H, s), 3.65 (1H, t, J = 8.7 Hz);
MS (ESI) m/z 460 [(M+H)+]; Rf 0.28 (MeOH/CHCl3
1:10); HRMS calcd for C29H50NO3 460.3791. Found
460.3794.
7.1.32.
N-Methyl-N-propyl-8-(17b-hydroxy-5a-andro-
stan-3-one-7a-yl)octanamide (20). This compound was
prepared from 18b using a procedure similar to that de-
scribed for 23. Yield: 58%. Colorless oil. 1H NMR
(270 MHz, CDCl3) d 0.76 (3H, s), 0.88 (3/2H, t,
J = 7.4 Hz), 0.92 (3/2H, t, J = 7.4 Hz), 1.01–1.90 (34H,
m), 1.04 (3H, s), 1.95–2.45 (8H, m), 2.91 (3/2H, s),
2.97 (3/2H, s), 3.23 (2/2H, t, J = 7.7 Hz), 3.33 (2/2H, t,
J = 7.7 Hz), 3.62–3.69 (1H, m); MS (ESI) m/z 488
[(M+H)+]; HRMS calcd for C31H54NO3 488.4104.
Found 488.4085.
7.1.27. N-n-Butyl-N-methyl-8-(17b-hydroxy-5a-andro-
stan-3-one-7a-yl)octanamide (19a). This compound was
prepared from 18b using a procedure similar to that de-
scribed for 23. Yield: 91%. Colorless oil. 1H NMR
(270 MHz, CDCl3) d 0.76 (3H, s), 0.92 (3/2H, t,
J = 7.4 Hz), 0.95 (3/2H, t, J = 7.4 Hz), 1.04 (3H, s),
1.18–2.45 (40H, m), 2.91 (3/2H, s), 2.97 (3/2H, s), 3.25
(2/2H, t, J = 7.6 Hz), 3.35 (2/2H, t, J = 7.6 Hz), 3.64
(1H, t, J = 8.2 Hz); MS (ESI) m/z 502 [(M+H)+]; HRMS
calcd for C32H56NO3 502.4260. Found 502.4239.
7.1.33. N-Isopropyl-N-methyl-8-(17b-hydroxy-5a-andro-
stan-3-one-7a-yl)octanamide (21). This compound was
prepared from 18b using a procedure similar to that
1
described for 23. Yield: 80%. Colorless oil. H NMR
(270 MHz, CDCl3) d 0.76 (3H, s), 0.90–1.82 (28H,
m), 1.04 (3H, s), 1.07 (6· 3/5H, d, J = 6.8 Hz), 1.18
(6· 2/5H, d, J = 6.6 Hz), 1.95–2.18 (3H, m), 2.20–
2.43 (5H, m), 2.77 (3· 2/5H, 2), 2.80 (3· 3/5H, s),
3.65 (1H, t, J = 8.8 Hz), 4.02–4.15 (1· 2/5H, m),
4.82–4.95 (1· 3/5H, m); MS (FAB) m/z 488
[(M+H)+]; HRMS calcd for C31H54NO3 488.4104.
Found 488.4099.
7.1.28. N-n-Butyl-N-methyl-9-(17b-hydroxy-5a-andro-
stan-3-one-7a-yl)nonanamide (19b). This compound was
prepared from 17c using a procedure similar to that de-
scribed for 23. Colorless oil. 1H NMR (270 MHz,
CDCl3) d 0.76 (3H, s), 0.92 (3/2H, t, J = 7.4 Hz), 0.95
(3/2H, t, J = 7.4 Hz), 1.04 (3H, s), 1.18–2.45 (42H, m),
2.91 (3/2H, s), 2.97 (3/2H, s), 3.25 (2/2H, t,
J = 7.6 Hz), 3.35 (2/2H, t, J = 7.6 Hz), 3.63 (1H, t,
J = 8.2 Hz); MS (ESI) m/z 538 [(M+Na)+]; HRMS calcd
for C33H58NO3 516.4417. Found 516.4391.
7.1.34. N-Methyl-N-ethyl-8-(17b-hydroxy-5a-androstan-
3-one-7a-yl)octanamide (22). This compound was pre-
pared from 18b using a procedure similarto that described
1
for 23. Yield: 72%. Colorless oil. H NMR (270 MHz,
CDCl3) d 0.76 (3H, s), 0.98–1.90 (32H, m), 1.04 (3H, s),
1.09 (3/2H, t, J = 7.1 Hz), 1.17 (3/2H, t, J = 7.1 Hz),
1.95–2.45 (8H, m), 2.91 (3/2H, s), 2.97 (3/2H, s), 3.34 (2/
2H, t, J = 7.1 Hz), 3.41 (2/2H, t, J = 7.7 Hz), 3.62–3.69
(1H, m); MS (ESI) m/z 474 [(M+H)+]; HRMS calcd for
C30H52NO3 474.3947. Found 474.3948.
7.1.29. N-n-Butyl-N-methyl-10-(17b-hydroxy-5a-andro-
stan-3-one-7a-yl)decanamide (19c). This compound
was prepared from 17d using a procedure similar to
that described for 23. Colorless oil. 1H NMR
(270 MHz, CDCl3) d 0.76 (3H, s), 0.92 (3/2H, t,
J = 7.4 Hz), 0.95 (3/2H, t, J = 7.4 Hz), 1.04 (3H, s),
1.18–2.45 (44H, m), 2.91 (3/2H, s), 2.97 (3/2H, s),
3.25 (2/2H, t, J = 7.4 Hz), 3.36 (2/2H, t, J = 7.4 Hz),
3.62–3.70 (1H, m); MS (ESI) m/z 530 [(M+H)+];
HRMS calcd for C34H60NO3 530.4573. Found
530.4575.
7.1.35. N,N-Diethyl-8-(17b-hydroxy-5a-androstan-3-one-
7a-yl)octanamide (24). This compound was prepared
from 18b using a procedure similar to that described
1
for 23. Yield: 95%. Colorless oil. H NMR (270 MHz,
CDCl3) d 0.76 (3H, s), 1.04 (3H, s), 1.00–1.83 (34H,
m), 1.95–2.16 (3H, m), 2.22–2.47 (5H, m), 3.30 (2H, q,
J = 7.1 Hz), 3.37 (2H, q, J = 7.1 Hz), 3.65 (1H, t,
J = 9.0 Hz); MS (ESI) m/z 488 [(M+H)+]; HRMS calcd
for C31H54NO3 488.4104. Found 488.4088.
7.1.30. N-n-Butyl-N-methyl-11-(17b-hydroxy-5a-andro-
stan- 3-one-7a-yl)undecanamide (19d). This compound
was prepared from 17e using a procedure similar to that
described for 23. Colorless oil. 1H NMR (270 MHz,
CDCl3) d 0.76 (3H, s), 0.92 (3/2H, t, J = 7.4 Hz), 0.95
(3/2H, t, J = 7.4 Hz), 1.04 (3H, s), 1.12–2.45 (46H, m),
2.91 (3/2H, s), 2.96 (3/2H, s), 3.25 (2/2H, t,
J = 7.4 Hz), 3.36 (2/2H, t, J = 7.4 Hz), 3.64 (1H, br);
MS (ESI) m/z 544 [(M+H)+]; HRMS calcd for
C35H62NO3 544.4730. Found 544.4728.
7.1.36.
N-Benzyl-N-methyl-8-(17b-hydroxy-5a-andro-
stan-3-one-7a-yl)octanamide (25). This compound was
prepared from 18b using a procedure similar to that de-
scribed for 23. Yield: 85%. Colorless oil. 1H NMR
(270 MHz, CDCl3) d 0.76 (3H, s), 0.97–1.82 (28H, m),
1.04 (3H, s), 1.95–2.15 (3H, m), 2.22–2.43 (5H, m),

K. Tachibana et al. / Bioorg. Med. Chem. 15 (2007) 174–185
183
2.92 (3· 4/7H, s), 2.94 (3· 3/7H, s), 3.60–3.70 (1H, m),
4.54 (2· 3/7H, s), 4.59 (2· 4/7H, s), 7.15–7.39 (5H, m);
MS (ESI) m/z 536 [(M+H)+]; HRMS calcd for
C35H54NO3 536.4104. Found 536.4103.
7.1.43. N,N-Dimethyl-7-(17b-hydroxy-5a-androstan-3-
one-7a-yl)heptanamide (36). This compound was pre-
pared from 7 using a procedure similar to that described
1
for 23. Yield: 76%. Colorless oil. H NMR (270 MHz,
CDCl3) d 0.76 (3H, s), 0.85–1.83 (26H, m), 1.04 (3H,
s), 1.95–2.15 (3H, m), 2.22–2.45 (5H, m), 2.94 (3H, s),
3.01 (3H, s), 3.65 (1H, t, J = 9.0 Hz); MS (FAB) m/z
446 [(M+H)+]; HRMS calcd for C28H48NO3 446.3634.
Found 446.3631.
7.1.37.
8-(17b-Hydroxy-5a-androstan-3-one-7a-yl)-1-
(morpholin-4-yl)octan-1-one (27). This compound was
prepared from 18b using a procedure similar to that de-
scribed for 23. Yield: 96%. Colorless oil. 1H NMR
(270 MHz, CDCl3) d 0.76 (3H, s), 0.97–1.82 (28H, m),
1.04 (3H, s), 1.95–2.15 (3H, m), 2.22–2.43 (5H, m),
3.45–3.48 (2H, m), 3.60–3.69 (7H, m); MS (ESI) m/z
502 [(M+H)+]; HRMS calcd for C31H52NO4 502.3896.
Found 502.3900.
7.1.44. N,N-Diethyl-7-(17b-hydroxy-5a-androstan-3-one-
7a-yl)heptanamide (37). This compound was prepared
from 7 using a procedure similar to that described for
23. Yield: 7.1%. Colorless oil. 1H NMR (270 MHz,
CDCl3) d 0.76 (3H, s), 0.82–1.90 (26H, m), 1.04 (3H,
s), 1.11 (3H, t, J = 7.2 Hz), 1.17 (3H, t, J = 7.2 Hz),
1.95–2.15 (3H, m), 2.20–2.45 (5H, m), 3.30 (2H, q,
J = 7.1 Hz), 3.37 (2H, q, J = 7.1 Hz), 3.65 (1H, t,
J = 9.0 Hz); MS (FAB) m/z 474 [(M+H)+]; HRMS calcd
for C30H52NO3 474.3947. Found 474.3954.
7.1.38.
8-(17b-Hydroxy-5a-androstan-3-one-7a-yl)-1-
(pyrrolidin-1-yl)octan-1-one (28). This compound was
prepared from 18b using a procedure similar to that de-
scribed for 23. Yield: 97%. Colorless oil. 1H NMR
(270 MHz, CDCl3) d 0.76 (3H, s), 0.97–2.10 (35H, m),
1.04 (3H, s), 2.22–2.43 (5H, m), 3.39–3.48 (4H, m),
3.61–3.70 (1H, m); MS (ESI) m/z 486 [(M+H)+]; HRMS
calcd for C31H52NO3 486.3947. Found 486.3956.
7.1.45.
one-7a-yl)octanamide (26). To
(10.3 mg, 0.0238 mmol) and O-(7-azabenzotriazol-1-
yl)-1,1,3,3-tetramethyluromium hexafluorophosphate
N,N-Dihexyl-8-(17b-hydroxy-5a-androstan-3-
a
solution of 18b
7.1.39.
8-(17b-Hydroxy-5a-androstan-3-one-7a-yl)-1-
i
(piperidin-1-yl)octan-1-one (29). This compound was
prepared from 18b using a procedure similar to that de-
scribed for 23. Yield: 88%. Colorless oil. 1H NMR
(270 MHz, CDCl3) d 0.76 (3H, s), 1.04 (3H, s), 1.00–
1.83 (34H, m), 1.95–2.16 (3H, m), 2.21–2.47 (5H, m),
3.39 (2H, t, J = 5.3 Hz), 3.54 (2 H, t, J = 5.3 Hz), 3.65
(1H, t, J = 9.0 Hz); MS (ESI) m/z 500 [(M+H)+]; HRMS
calcd for C32H54NO3 500.4104. Found 500.4104.
(30 mg, 0.0789 mmol) in THF (1 ml) were added Pr2-
NEt (25 ll, 0.0167 mmol) and di-n-hexylamine (17.2 ll,
0.0735 mmol), and the mixture was stirred at room tem-
perature for 2 h. After adding AcOEt, the reaction mix-
ture was washed with satd NaHCO3 aq, 1 N HCl, and
brine. To the organic layer, NH silica gel (Pro. No.
DM1020; Fuji Silicia Chemical Co., Ltd) was added
and the mixture was stirred for 5 min. After filtering,
the solvent was distilled off at reduced pressure. The
resulting residue was purified by silica gel column chro-
matography (MeOH/CHCl3 1:10) to give 26 (9.5 mg,
7.1.40. N-Methyl-8-(17b-hydroxy-5a-androstan-3-one-
7a-yl)octanamide (31). This compound was prepared
from 18b using a procedure similar to that described
1
67%) as colorless oil. H NMR (270 MHz, CDCl3) d
1
for 23. Yield: 72%. Colorless oil. H NMR (270 MHz,
0.76 (3H, s), 0.88 (3H, t, J = 6.9 Hz), 0.90 (3H, t,
J = 6.9 Hz), 1.04 (3H, s), 1.00–1.83 (44H, m), 1.95–
2.16 (3H, m), 2.21–2.47 (5H, m), 3.19 (2H, t,
J = 8.5 Hz), 3.28 (2H, t, J = 8.5 Hz), 3.65 (1 H, t,
J = 9.0 Hz); MS (ESI) m/z 600 [(M+H)+]; HRMS calcd
for C39H70NO3 600.5356. Found 600.5332.
CDCl3) d 0.76 (3H, s), 0.97–1.85 (28H, m), 1.04 (3H,
s), 1.98–2.42 (8H, m), 2.81 (3H, d, J = 4.9 Hz), 3.61–
3.70 (1H, m), 5.48 (1H, br); MS (ESI) m/z 446
[(M+H)+]; HRMS calcd for C28H48NO3 446.3634.
Found 446.3655.
7.1.41. N-Ethyl-8-(17b-hydroxy-5a-androstan-3-one-7a-
yl)octanamide (32). This compound was prepared from
18b using a procedure similar to that described for 23.
7.1.46. N-Propyl-8-(17b-hydroxy-5a-androstan-3-one-7a-
yl)octanamide (33). This compound was prepared from
18b using a procedure similar to that described for 26.
1
1
Yield: 64%. Colorless oil. H NMR (270 MHz, CDCl3)
Yield: 99%. Colorless oil. H NMR (270 MHz, CDCl3)
d 0.76 (3H, s), 0.97–1.85 (28H, m), 1.04 (3H, s), 1.14
(3H, t, J = 7.3 Hz), 1.98–2.42 (8H, m), 3.24–3.35 (2H,
m), 3.65 (1H, t, J = 8.2 Hz), 5.43 (1H, br); MS (FAB)
m/z 460 [(M+H)+]; HRMS calcd for C29H50NO3
460.3791. Found 460.3811.
d 0.76 (3H, s), 0.92 (3H, t, J = 7.6 Hz), 0.95–1.82
(30 H, m), 1.04 (3H, s), 1.93–2.45 (8H, m), 3.21 (2H,
dt, J = 6.3, 6.8 Hz), 3.65 (1H, t, J = 9.0 Hz), 5.48 (1H,
br); MS (FAB) m/z 474 [(M+H)+]; HRMS calcd for
C30H52NO3 474.3947. Found 474.3938.
7.1.42. N-Butyl-8-(17b-hydroxy-5a-androstan-3-one-7a-
yl)octanamide (34). This compound was prepared from
18b using a procedure similar to that described for 23.
7.1.47. N-Hexyl-8-(17b-hydroxy-5a-androstan-3-one-7a-
yl)octanamide (35). This compound was prepared from
18b using a procedure similar to that described for 26.
1
1
Yield: 92%. Colorless oil. H NMR (270 MHz, CDCl3)
Yield: 99%. Colorless oil. H NMR (270 MHz, CDCl3)
d 0.76 (3H, s), 0.92 (3H, t, J = 7.3 Hz), 0.95–1.82
(32H, m), 1.04 (3H, s), 1.93–2.45 (8H, m), 3.25 (2H,
dt, J = 6.1, 7.1 Hz), 3.65 (1H, t, J = 9.0 Hz), 5.44 (1H,
br); MS (ESI) m/z 488 [(M+H)+]; HRMS calcd for
C31H54NO3 488.4104. Found 488.4099.
d 0.76 (3H, s), 0.88 (3H, t, J = 6.9 Hz), 1.04 (3H, s),
1.10–1.82 (36H, m), 1.93–2.45 (8H, m), 3.24 (2H, dt,
J = 6.1, 6.9 Hz), 3.65 (1H, t, J = 9.0 Hz), 5.45 (1H, br);
MS (ESI) m/z 516 [(M+H)+]; HRMS calcd for
C33H58NO3 516.4417. Found 516.4399.

184
K. Tachibana et al. / Bioorg. Med. Chem. 15 (2007) 174–185
7.1.48. 8-(17b-Hydroxy-5a-androstan-3-one-7a-yl)octana-
mide (30). To a solution of 18b (12.6 mg, 0.0291 mmol)
in THF (0.5 ml) were added NEt3 (8.1 ll, 0.0584 mmol)
and ClCOOEt (4.2 ll, 0.0441 mmol) dropwise at 0 ꢂC.
After stirring for 5 min, ammonia gas was bubbled into
the solution for 30 s. After stirring for 30 min, water was
added to the reaction mixture and extracted with
CH2Cl2. The organic layer was washed with brine, dried
over MgSO4, filtered, and concentrated at reduced pres-
sure. Purification by silica gel column chromatography
(CH2Cl2/MeOH 20:1) gave 30 (11.6 mg, 92%) as color-
Twenty-four hours before transfection, 1.0 · 105 HeLa
cells were cultured in phenol red-free DMEM/5%
DCC-FBS on 12-well microplates. Five hundred nano-
grams/well of MMTV-Luc vector, 100 ng/well of
pSG5-hAR, and 5 ng/well of Renilla Luc vector were
transfected into the HeLa cells. The transfection was
performed in a liquid culture of the phenol red-free
DMEM using 3 ml/well of Lipofectamine. Nine hours
after the transfection, the liquid culture was replaced
by phenol red-free DMEM/3% DCC-FBS containing
0.1 nmol/L of DHT and 1, 10, 100, 1000, or
10,000 nmol/L of test compound. The transcriptional
activity value was measured 48 h after the replacement
of the liquid culture. Transcriptional activity was
measured with a dual luciferase reporter assay system.
The transcriptional activity value was calculated as the
value for firefly luciferase divided by the value for
sea pansy luciferase. The antagonist activity was com-
puted by the following formula and the determined
antagonist activity was used to compute the IC50 value
(the concentration for a compound-treated group at
which it shows a 50% decrease in the transcriptional
activity of DHT 0.1 nmol/L when the compound was
not added).
1
less oil. H NMR (270 MHz, CDCl3) d 0.76 (3H, s),
0.98–1.83 (28H, m), 1.04 (3H, s), 1.95–2.16 (3H, m),
2.22–2.47 (5H, m), 3.65 (1H, t, J = 8.3 Hz), 5.41 (2H,
br); MS (FAB) m/z 432 [(M+H)+]; HRMS calcd for
C27H46NO3 432.3478. Found 432.3498.
7.2. Biological assay
7.2.1. Competitive binding assay. CHO-K1/hAR cells
(5 · 104/well) were plated in 24-well plates and cultured
for 2 days. Adhered cells were washed with PBS(ꢀ) and
replaced with phenol red-free DMEM containing
0.34 nmol/L [3H]mibolerone in the presence or absence
of test compound. Nonspecific binding of [3H]miboler-
one was determined separately by adding 200-fold ex-
cess of cold mibolerone. Following 2 h incubation at
37 ꢂC, cells were washed with PBS(ꢀ) and solubilized
in 10 mmol/L Tris–HCl, pH 6.8, containing 2% SDS
and 10% glycerol. Radioactivity was counted using a
scintillation counter.
Antagonist activity ð%Þ
¼ðCompound-treated transcriptional activity=
Non-treated transcriptional activityÞꢁ100:
Acknowledgments
The binding affinity was described as an IC50 value (the
concentration of a compound required to inhibit 50% of
[3H]mibolerone).
The authors thank Mr. Kenichiro Kotake and cowork-
ers for HRMS measurements of the compounds, and
Ms. Frances Ford of Chugai Pharmaceutical Co., Ltd
for assistance with English usage.
7.2.2. Reporter gene assay. Twenty-four hours before
transfection, 1.0 · 105 HeLa cells were cultured in phe-
nol red-free DMEM/5% DCC-FBS on 12-well micro-
plates. Five hundred nanograms/well of MMTV-Luc
vector, 100 ng/well pSG5-hAR, and 5 ng/well of Renilla
Luc vector were transfected into the HeLa cells. The
transfection was performed in a liquid culture of the
phenol red-free DMEM using 3 ml/well of Lipofect-
amine. Nine hours after the transfection, the liquid cul-
ture was replaced by phenol red-free DMEM/3% DCC-
FBS containing 1, 10, 100, 1000, or 10,000 nmol/L of
test compound. The transcriptional activity value was
measured 48 h after the replacement of the liquid cul-
ture. Transcriptional activity was measured with a dual
luciferase reporter assay system. The transcriptional
activity value was calculated as the value for firefly lucif-
erase divided by the value for sea pansy luciferase. The
agonist activity was computed by the following formula
and the determined agonist activity was used to compute
the FI5 value (the concentration of a compound-treated
group that shows a transcriptional activity five times the
transcriptional activity of the group without a test
compound).
References and notes
1. Landis, S. H.; Murray, T.; Bolden, S.; Wingo, P. A. CA
Cancer J. Clin. 1999, 49, 8.
2. Stewart, A. B.; Lwaleed, B. A.; Douglas, D. A.; Birch, B.
R. Curr. Med. Chem. Anti-canc. Agents 2005, 5, 603.
3. Kennealey, G. T.; Furr, B. J. Urol. Clin. North Am. 1991,
18, 99.
4. Blackledge, G.; Kolvenbag, G.; Nash, A. Anticancer
Drugs 1996, 7, 27.
5. Tyrrell, C. J.; Altwein, J. E.; Klippel, F.; Varenhorst, E.;
Lunglmayr, G.; Boccardo, F.; Holdaway, I. M.; Haefliger,
J. M.; Jordaan, J. P. J. Urol. 1991, 146, 1321.
6. Eisenberger, M. A.; Blumenstein, B. A.; Crawfold, E. D.;
Miller, G.; McLeod, D. G.; Leohrer, P. J.; Wilding, G.;
Sears, K.; Culkin, D. J.; Thompson, I. M., Jr.; Bueschen,
A. J.; Lowe, B. A. N. Engl. J. Med. 1998, 339, 1036.
7. Wirth, M. P.; Froschermaier, S. E. Urol. Res. 1997,
25(Suppl. 2), S67.
8. Kemppainen, J. A.; Wilson, E. M. Urology 1996, 48, 157.
9. Brunner, N.; Frandsen, T. L.; Holst-Hansen, C.; Bei, M.;
Thompson, E. W.; Wakeling, A. E.; Lippman, M. E.;
Clarke, R. Cancer Res. 1993, 53, 3229.
Agonist activity¼Compound-treated transcriptional activity=
Non-treated transcriptional activity:
10. Bowler, J.; Lilley, T. J.; Pittam, J. D.; Wakeling, A. E.
Steroids 1989, 54, 71.

K. Tachibana et al. / Bioorg. Med. Chem. 15 (2007) 174–185
185
11. Osborne, C. K.; Pippen, J.; Jones, S. E.; Parker, L. M.;
Ellis, M.; Come, S.; Gertler, S. Z.; May, J. T.; Burton, G.;
Dimery, I.; Webster, A.; Morris, C.; Elledge, R.; Buzdar,
A. J. Clin. Oncol. 2002, 20, 3386.
12. Howell, A.; Robertson, J. F. R.; Quaresma Albano, J.;
Aschermannova, A.; Mauriac, L.; Kleeberg, U. R.;
Vergote, I.; Erikstein, B.; Webster, A.; Morris, C. J. Clin.
Oncol. 2002, 20, 3396.
17. Brzozowski, A. M.; Pike, A. C. W.; Dauter, Z.; Hubbard,
¨
R. E.; Bonn, T.; Engstro¨m, O.; Ohman, L.; Greene, G. L.;
˚
Gustafsson, J. A.; Carlquist, M. Nature 1997, 389, 753.
18. He, W.; Huang, F.-C.; Morytko, M.; Jariwala, N.; Yu, K.-
T. Bioorg. Med. Chem. Lett. 1998, 8, 2825.
19. O’Reilly, J. M.; Li, N.; Duax, W. L.; Brueggemeier, R. W.
J. Med. Chem. 1995, 38, 2842.
20. Nickisch, K.; Laurent, H. Tetrahedron Lett. 1988, 29, 1533.
13. Pike, A. C. W.; Brzozowski, A. M.; Walton, J.; Hubbard,
˚
R. E.; Thorsell, A. G.; Li, Y. L.; Gustafsson, J. A.;
21. Brummer, O.; Ruckert, A.; Blechert, S. Chem. Eur. J.
¨
1997, 3, 441.
¨
Carlquist, M. Structure 2001, 9, 145.
14. Ishioka, T.; Tanatani, A.; Nagasawa, K.; Hashimoto, Y.
Bioorg. Med. Chem. Lett. 2003, 13, 2655.
15. Hashimoto, Y.; Miyachi, H. Bioorg. Med. Chem. 2005, 13,
5080.
16. Matias, P. M.; Donner, P.; Coelho, R.; Thomaz, M.;
Peixoto, C.; Macedo, S.; Otto, N.; Joschko, S.; Scholz, P.;
Wegg, A.; Ba¨sler, S.; Scha¨fer, M.; Egner, U.; Carrondo,
M. A. J. Biol. Chem. 2000, 275, 26164.
22. Kanbe, Y.; Kim, M.-H.; Nishimoto, M.; Ohtake, Y.;
Kato, N.; Tsunenari, T.; Taniguchi, K.; Ohizumi, I.;
Kaiho, S.; Morikawa, K.; Jo, J.-C.; Lim, H.-S.; Kim, H.-
Y. Bioorg. Med. Chem. 2006, 14, 4803.
23. Wong, C.-I.; Kelce, W. R.; Sar, M.; Wilson, E. M. J. Biol.
Chem. 1995, 270, 19998.
24. Hollenberg, S. M.; Evans, R. M. Cell 1988, 55, 899.
25. Simental, J. A.; Sar, M.; Lane, M. V.; French, F. S.;
Wilson, E. M. J. Biol. Chem. 1991, 266, 510.